METHODS

We identified all relevant published articles or conference abstracts relating to the epidemiology, natural history, and treatment of HIV-1 and HCV coinfection over the past nine years. Medline and AIDSLINE databases were first searched using the terms “HIV” and “HCV”, “liver disease”, and for the treatment section we also used the terms “interferon”, “antiretroviral therapy” in combination. Abstracts were included only where complete data were available. For prevalence studies we only included studies with ≥100 patients. In the section on perinatal transmission, studies were included if they had at least 20 coinfected mothers and used either the presence of positive hepatitis C RNA at any time or HCV antibody detection at 12 months for the diagnosis of HCV infection in newborn babies. For the impact of HIV-1 on HCV progression to fibrosis, cirrhosis, decompensated liver disease, or liver related deaths, we included all studies that investigated the effect of coinfection on any aspect of progression or mortality, including HCV viral load and transaminases. The same criteria were applied to studies examining the impact of HCV on HIV progression to AIDS or death, and on the impact of HCV treatment on HIV and HCV disease progression.

PREVALENCE

We identified 12 published seroprevalence studies based on various HIV-1 infected cohorts from Europe^10–,¹³ and North America,^14–,²¹ as shown in table 1⇓. HCV prevalence ranged between 7% and 57% and these differences in seroprevalence across the different studies were largely determined by the HCV risk factor distribution of the study population. Patients with a current or previous history of drug use had infection rates in excess of 80%^{12–14,16,17,20} while the prevalence of HCV in homo/bisexual groups varied between 2.6% and 15.2%.^{10,12,13,15–17,20} The majority (98%) of HIV-1 infected haemophiliacs are also coinfected with hepatitis C.^6,22 However, even these data may underestimate the true prevalence of HCV among HIV positive patients as at least 4% of HIV-HCV coinfected patients have no detectable antibodies in the presence of HCV viraemia,^24,25 or as a result of immunosuppression may subsequently lose detectable antibodies from serum despite persistent viraemia.²⁶

“The majority (98%) of HIV-1 infected haemophiliacs are also coinfected with hepatitis C”

The consistently high prevalence of HCV infection observed in HIV-1 infected individuals supports the routine screening for HCV in these patients, especially among haemophiliacs and drug users. Where HCV is suspected in the setting of negative antibody screening, detection of HCV RNA by polymerase chain reaction is recommended.

TRANSMISSION

HIV and HCV viruses share common routes of transmission, especially the parenteral route, and as a result coinfection rates in intravenous drug users and haemophiliacs are particularly high (60–90%).^6,22,23 Other non-parenteral routes of transmission are also important,^2,27,28 and there is now increasing evidence that sexual^15,27,29,30 and mother to child HCV transmission is facilitated by HIV infection.^18,31–,³⁵ In a study of 662 homosexual men, patients with HIV infection were three times more likely to be HCV positive compared with those who were HIV negative (9% v 3%; p<0.001).³⁰ In addition, independent sexual transmission of HCV among homosexual men accounted for approximately 50% of these cases.³⁰

“There is now increasing evidence that sexual^15,27,29,30 and mother to child HCV transmission is facilitated by HIV infection”

Similarly, there is a two to threefold higher maternal fetal transmission rate in coinfected mothers compared with a transmission rate of less than 6% in HIV negative mothers.^18,31–,³⁹

Table 2⇓ summarises 10 studies of mother to child transmission in coinfected women comprising 1089 coinfected mothers. The vertical transmission rate ranged from a low of 5.8% in one study from the UK to 36% in a large study from Italy. Several factors were associated with a higher rate of HCV transmission in coinfected patients. These included a higher HCV viral load,^18,31 vaginal delivery,^32,34,35 and in one study breastfeeding increased HCV transmission by fourfold (p=0.002) even after adjusting for mode of delivery.³⁵ In contrast, caesarean section reduced the rate of HCV transmission,^32,34,35 although this has not been corroborated in studies of HIV uninfected/HCV positive mothers, and cannot currently be recommended for management.³⁵ Few studies have addressed the converse issue of the impact of HCV on HIV transmission in coinfected mothers. However, one study of 487 HIV infected mothers of whom 161 were also HCV infected found a 1.82-fold increased rate of mother to child HIV transmission (95% confidence interval (CI) 1.12–2.95) in coinfected mothers.⁴⁰

The majority of studies have reported higher HCV viral loads by a magnitude of 0.3–1.08 log RNA copies/ml in coinfected patients^25,41–,⁵³ compared with HIV-1 negative HCV infected subjects (table 3⇓), and this is likely to mediate the higher transmission of HCV in the setting of HIV infection. However, one report suggests that this may be the case only in patients infected with HCV genotype 1.⁵⁴ An association between serum HCV viral load and saliva HCV viral load has also been recently demonstrated, which may have major implications for HCV transmission in coinfected patients.⁵⁵

Together, these data demonstrate a higher rate of vertical and sexual transmission of HCV in coinfected patients, and that HIV is a cofactor for HCV transmission.

IMPACT OF HIV ON HCV PROGRESSION

Prior to HAART, the majority of deaths in HIV infected patients were AIDS related, and data on HCV related morbidity and mortality were lacking.¹⁵ However, since HAART became widely available in the last five years and extended life expectancy, the management of concurrent illnesses has attracted more attention. Table 4⇓ summarises data from 17 studies^{5,6,15,20,46,53,56,66} on the impact of HIV infection on HCV progression, comprising a total of 2509 coinfected patients. A recent meta-analysis included eight of these studies^56–,⁶³ and showed a combined adjusted relative risk of 2.92 (95% CI 1.70–5.0) for progression to cirrhosis or decompensated liver disease in coinfected patients.⁶⁷ Similar evidence for liver related deaths was provided by the UK haemophiliac cohort of 305 patients of whom 125 were coinfected. The relative hazard of liver related death, adjusted for age and HCV genotype, was 17.5-fold for coinfected patients (95% CI 5.8–52.7) compared with those infected with HCV alone.⁶ More recently, the UK National Haemophilia Register has reported that men with haemophilia exposed to HCV and HIV are 4.6-fold more likely to progress to liver related death (6.5%; range 4.5–9.5%) compared with HIV-1 uninfected men (1.4%; range 4.5–9.5%).⁵

“Men with haemophilia exposed to HCV and HIV are 4.6-fold more likely to progress to liver related death compared with HIV-1 uninfected men”

Identification of factors influencing HCV disease progression in HIV coinfected patients has been based mainly on haemophiliac cohorts and has not been assessed adequately in other HIV risk groups. Level of CD4 immunosuppression has emerged as one of the most important determinants of progression to liver fibrosis, and patients with CD4 cell counts less than 500 cells×10⁹/l are 3.2 times (95% CI 1.1–9.2) more likely to have advanced liver fibrosis on liver biopsy.⁶⁴ Patients with a low CD4 count or who had an AIDS diagnosis were also at increased risk for severe liver disease.^59–61,64 Other host factors associated with more rapid HCV disease progression include older age at infection and excess alcohol intake.⁵⁹ As for HCV transmission, the impact of HIV on HCV progression is likely to be partly mediated through the increased HCV viral load (table 3⇑). A higher HCV viral load was shown in three studies to correlate with increased hepatic inflammation, as defined by a higher alanine aminotransferase (ALT) value^43,49,50 but this has not been confirmed in other studies.^{25,41,42,45,46,48,51,54} In addition, an association between HCV viral load and fibrosis progression has not been clearly demonstrated even in HIV-1 uninfected HCV patients, and a positive correlation with histological grade and stage of disease is lacking.⁶⁸ There are conflicting data on the impact of HCV genotype on disease progression in coinfected patients. Two studies have shown no effect of genotype on progression^57,64 while a further two have shown the presence of genotype 1 to be associated with more advanced fibrosis stage⁶⁵ and liver related death.⁶ Other potential risk factors such as mode of acquisition, past hepatitis B infection, and HCV viral load have not been assessed in coinfected populations.

HCV is a known carcinogen,⁶⁹ and the rising prevalence of infection has been implicated as a major contributing factor to the recent increase in incidence of hepatocellular carcinoma (HCC) in the USA.⁷⁰ The precise mechanism by which HCV induces carcinogenesis is unclear, but animal studies in transgenic mice have implicated the core protein,⁷¹ which has been reported to downregulate p53 promoter activity⁷² and repress p21 promoter activity through the p53 pathway.⁷³ These events may form the basis for the direct role of HCV in the induction of HCC. It has recently been reported that HCC occurs at a younger age in coinfected patients compared with those infected with HCV alone.⁷⁴ While further follow up data are needed, this trend is expected to continue over the coming years as the impact of HCV related liver injury in HIV infected subjects gains further attention.

“HCC occurs at a younger age in coinfected patients compared with those infected with HCV alone”

The consensus from these studies is that HIV-1 is clearly associated with accelerated liver disease and reduced survival in HCV infected patients. In altering the rate of fibrosis progression, HIV-1 may further alter the natural history of HCV infection resulting in an aggressive course to end stage liver disease and liver failure.

IMPACT OF HCV ON HIV-1 PROGRESSION

The impact of HCV infection on the course of HIV disease has become evident in recent years, and the findings of nine studies are summarised in table 5⇓,^{7,12,15,20,21,58,75–},⁷⁷ four of which were conducted before the widespread availability of triple combination of antiretroviral therapy. In five studies HCV infection appears to have a significant effect on the progression of HIV to AIDS defining illness and AIDS related mortality.^{7,12,58,75,76} In the largest study reported to date, based on a Swiss HIV cohort of patients with well controlled HIV-1 replication, 1593 of whom were coinfected, the risk of progression to AIDS defining illness or death was 3.54 (95% CI 2.0–6.25) compared with HCV uninfected individuals. Daar et al have also reported a detrimental effect of HCV viral load on HIV progression. For every 10-fold increase in baseline HCV viral load, the relative risk for clinical progression to AIDS was 1.66 (95% CI 1.1–2.51), and the relative risk for AIDS related mortality was 1.54 (95% CI 1–2.3), even after controlling for CD4⁺ cell count and HIV-1 RNA level.⁷⁵ The evidence from these studies confirms that HCV is an independent factor associated with HIV progression to AIDS and AIDS related death. However, four studies failed to demonstrate a negative effect of HCV on HIV progression.^15,20,21,77 In one study, 70% of patients already had AIDS²⁰; two studies were based on retrospective¹⁵ or cross sectional analyses,²¹ and in a further study the follow up was 30 months.⁷⁷ Reasons for the adverse effect of HCV on HIV disease are unknown. One possible explanation is that patients with coinfection, the majority of whom have acquired their infection through injection drug use, may have reduced compliance with HIV therapies.

“HCV is an independent factor associated with HIV progression to AIDS and AIDS related death”

Thirteen studies have examined the impact of HCV viral load on CD4 cell counts (table 6⇓); eight of these have highlighted an inverse correlation between HCV viral load and CD4 count^{41,43,46,47,49,50,62} while five studies have shown no relationship,^{25,42,44,48,54} and one study reported a positive correlation.⁴⁵ It has been suggested that this inverse relationship between HCV viral load and CD4 count may be explained on the basis of immune dysregulation of HCV replication.^49,75

PATHOGENESIS OF HIV-1-HCV INTERACTION

Several lines of evidence suggest that liver injury in HCV patients is due to the immune response to HCV rather than a direct viral effect. Strong HCV specific T helper 1 cell responses and T helper cell recognition of multiple core epitopes are associated with clearance of HCV naturally and after interferon therapy.^78–,⁸⁰ Furthermore, HCV specific CD4+ and CD8+ T cells have been shown to persist as biomarkers for prior HCV exposure and recovery, even if HCV antibodies decline and become undetectable.⁸¹ HIV-1 infects CD4 cells which leads to impaired response of both the infected CD4+ T cells and the uninfected CD8+ T cells, and this functional loss eventually leads to profound immune dysregulation.⁸² The impaired immune response to HCV in the setting of HIV infection may explain the inability of coinfected patients to clear HCV infection naturally, and possibly also the poor response to current treatment.

“The mechanism by which HCV influences HIV-1 progression remains speculative”

The mechanism by which HCV influences HIV-1 progression remains speculative. HCV may downregulate proliferation of T cells⁸³ or increase apoptosis of T cells by apoptotic pathways.⁸⁴ Patients with HCV infection express Fas on peripheral blood mononuclear cells and HCV-RNA has preferentially been detected in these Fas positive cells.⁸⁵ This may form the basis for a synergistic effect of HIV-1 and HCV on CD4 positive cells both in terms of production and apoptosis which in turn could explain the negative impact of HCV on HIV progression.

IMPACT OF HIGHLY ACTIVE ANTIRETROVIRAL HIV THERAPY ON HCV PROGRESSION

A total of eight studies examined the impact of HAART on HCV progression. Five studies found no evidence for an effect of HAART on HCV replication^86–,⁹⁰ while two studies have reported significant transient increases in HCV viral load^91,92 and transaminases (ALT, aspartate aminotransferase (AST)) following introduction of HAART. A protective effect of HAART on fibrosis progression has also been suggested⁵⁹ but needs to be confirmed in large prospective studies. One report has also shown a significant increase in HCV viral load at 96 weeks post HAART compared with baseline, although samples taken at 24 weeks post HAART did not show a significant increase.⁹³ The mechanism for the higher HCV-RNA levels with HAART despite immune restoration is unclear but is probably not associated with HIV related immune dysfunction. However, the increase in transaminases is associated with an improving immune response as shown by a correlation with a reduction in HIV viral load and increase in CD4 cell count.^90–92,94 In one study, multivariate logistic regression analysis demonstrated that a HAART induced CD4 cell count increase of 50 cells or more was independently associated with severe hepatotoxicity (odds ratio 3.6; 95% CI 1.0–12.9) in coinfected patients.⁹⁴

A number of studies have also identified HCV as a strong predictor of the development of hepatotoxicity (elevation of AST and ALT of over 1.25 times from baseline) with HAART,^13,94–,⁹⁷ in addition to other factors such as ritonavir use^94,97 and hyperbilirubinaemia associated with indinavir treatment. Of 138 HIV patients treated with indinavir, hyperbilirubinaemia occurred in 56 (40.6%), and HCV was an independent predictor of hepatotoxicity.⁹⁸ Hepatotoxicity has also been observed with the use of reverse transcriptase inhibitors which can cause mitochondrial toxicity and microsteatosis, and which may also lead to lethal hepatotoxicity.⁹⁹ More recently, HCV was shown to be a confounding factor for the development of lipoatrophy in coinfected patients (p=0.003), and in a multivariate analysis HCV was independently associated with insulin resistance, body mass index, and peripheral fat wasting.¹⁰⁰

In conclusion, there is no definitive evidence to support a clear effect of HAART on the natural history of HCV infection but HCV infection appears to be an independent predictor of hepatotoxicity following the introduction of HAART. The relationship of liver toxicity to antiretroviral therapy depends on the specific antiretroviral agent and the degree of immune reconstitution.

IMPACT OF HCV TREATMENT ON HIV DISEASE PROGRESSION

Treatment of HCV infection has evolved over the last decade with an increasingly higher percentage of patients achieving sustained viral clearance (defined as HCV RNA negative at six months after stopping therapy) following the introduction of combination based therapies with interferon and ribavirin, and more recently pegylated interferon.^101–,¹⁰⁴ Of note, HCV-HIV coinfected patients were all excluded from placebo controlled trials with interferon alpha. However, the results of nine non-placebo controlled trials^6,52,105–,¹¹¹ showed an equivalent percentage who had a sustained response to HCV therapy in coinfected compared with HCV only infected individuals. This ranged between 8% and 29%, although the interferon dosage and duration varied considerably across the different studies (from 3 to 9 million units weekly for 6–12 months). Side effects of interferon in coinfected patients were comparable with HIV uninfected patients.⁵² Over 90% of patients were able to complete their treatment course indicating that interferon is well tolerated even when used in conjunction with antiretroviral therapy. A 5% drop in CD4 cell count occurred in less than 5% of patients treated, and occurred in the first 10 weeks of therapy. However, this decline in CD4 count may be clinically significant and contribute to the subsequent development of opportunistic infections.^52,112,113

Combination therapy with interferon alpha and ribavirin is the current gold standard in the treatment of HCV infection.^103,104 Based on three studies, 80 coinfected patients treated with interferon and ribavirin showed a comparable sustained viral response rate and incidence of side effects to HIV uninfected patients,^114–,¹¹⁶ and the results of several large studies are awaited. Five cases of mitochondrial toxicity have been reported in coinfected patients receiving interferon and ribavirin and were thought to relate to the interaction of ribavirin with didanosine^117,118 as a result of enhanced phosphorylation of didanosine which reached toxic levels.¹¹⁹ In all studies published thus far, the HIV viral load did not change when ribavirin was commenced in patients who were already receiving antiretroviral therapy.

“The development of a helicase inhibitor, which is anticipated in the near future, together with other novel approaches, are antisense genes, ribozymes, and HCV specific protease inhibitors”

The recent introduction of pegylated interferon in combination with ribavirin appears to further improve the rate of sustained viral response and patient compliance. Preliminary data report that the regimen is well tolerated with 65% achieving viral clearance at six months.¹²⁰ The development of a helicase inhibitor which is anticipated in the near future, together with other novel approaches, are antisense genes, ribozymes, and HCV specific protease inhibitors. Immunotherapy may also have a place in the future. Schlaak et al showed HCV viral clearance in two (28.6%) of seven coinfected patients treated with interleukin 2.¹²¹ Future HCV therapy is therefore likely to include multiple combinations similar to those used in HIV therapy.

TRANSPLANTATION

In the past, the presence of HIV infection was generally considered a contraindication for liver transplantation. This was partly related to the ethical debate over allocation of limited health care resources and the difficulty in justifying the use of live related donors when life expectancy was limited for other reasons. However, following the introduction of HAART and subsequent prolonged survival of HIV patients, organ transplantation as a treatment strategy for HCV infection is being evaluated in a number of centres. At King's College Hospital, 1000 liver recipient transplants were performed between 1995 and 2001, of which 10 patients were HIV infected¹²² and five had HIV-1-HCV coinfection. Three presented with fulminant liver failure, in two due to hepatitis B infection, in one due to non-A non-B infection (although he was also infected with hepatitis B), and in one due to end stage alcoholic liver disease. All survived the postoperative period and were discharged home. Four patients died subsequently at 3, 6, 15, and 25 months following transplantation, and the cause of death in these patients was hepatitis C complications as a result of reinfection.

“Organ transplantation as a treatment strategy for HCV infection is being evaluated in a number of centres”

None of the patients with hepatitis B developed graft reinfection post transplant, and six patients are currently alive with the longest survival now over 42 months. Only one patient developed complications related to HIV infection. The treatment related immunosuppressive drugs were well tolerated and infective complications and HIV progression were uncommon. Thus liver transplantation may be an option in carefully selected patients.

INTERACTION OF HCV-HIV-1 COINFECTION AND HEPATITIS A AND HEPATITIS B

There are no available data on vaccination in coinfected patients and therefore vaccination of HCV-HIV infected patients remains controversial. Acute hepatitis A has been shown to be more severe and associated with higher mortality in patients infected with HCV and especially in those with advanced liver disease.¹²³ However, other groups have not confirmed this finding.¹²⁴ The response of hepatitis A vaccination in HIV positive patients is high, with up to 88% mounting an antibody response and without adverse effect.¹²⁵

“Vaccination of HCV-HIV infected patients remains controversial”

Selective hepatitis A vaccination may be indicated in HIV-HCV coinfected patients who have not been exposed previously. Past hepatitis B exposure as defined by hepatitis B core antibody positive in HIV uninfected patients has also been shown to be a strong predictor of severe liver disease.² This observation was confirmed by paired liver biopsies where past hepatitis B infection was an independent marker of progression. Furthermore, hepatitis B is associated with more advanced liver disease in patients with chronic hepatitis C.¹²⁶ Patients with HIV-HCV coinfection who have not been exposed to hepatitis B should therefore be vaccinated. Currently, combined hepatitis A and B vaccination is available and will facilitate more convenient delivery of protective vaccination.